Research On Disaster Discrimination And Public Opinion Of Rainstorm Waterlogging Based On Weibo Text

Rainstorm waterlogging disasters refer to meteorological disasters that cause serious water accumulation in urban areas and affect public safety due to the rapid and heavy rainstorm and the poor drainage of cities.Its main characteristics are sudden,group,continuous and frequent,causing huge economic losses.With the acceleration of urbanization in China,how to prevent and reduce economic losses and casualties caused by rainstorm and waterlogging has become a hot and difficult issue.At the same time,the timeliness and accuracy of traditional media channels are low,and microblog as a new social media has a strong timeliness.Therefore,there is an urgent need for a micro-blog text-based rainstorm waterlogging disaster identification and public opinion research methods to deal with.In this paper,a method based on micro-blog text to distinguish rainstorm waterlogging disaster is proposed.As a new social media,microblog has strong instantaneity and a large number of users.Therefore,it is of theoretical and practical significance to use text categorization algorithm to discriminate rainstorm waterlogging based on micro-blog text.Firstly,this paper obtains the microblog data after rainstorm and waterlogging disasters,uses three text classification algorithms to classify the text of microblog,and makes a binary classification judgment on whether the text of microblog is disaster-prone or not.Based on three commonly used performance evaluation indexes of text classification algorithm,the classification performance of three text classification algorithms is compared.The experiment shows that Naive Bayesian algorithm has the best classification performance.It is applied to the practice of identifying the flood situation of rainstorm based on micro-blog text.On the basis of realizing the identification of rainstorm waterlogging disaster,this paper also puts forward a new idea of Public Opinion Research on Rainstorm Waterlogging Disaster Based on micro-blog text.After the rainstorm waterlogging disaster occurs,the text categorization algorithm is used to identify the disaster situation of the text data obtained from the microblog,and the public opinion of the text identified as disaster situation is studied.The rainstorm waterlogging disaster and public sentiment are visualized,so that the relevant departments can carry out targeted emergency management and rescue,in order to reduce casualties and economic losses caused by the rainstorm waterlogging disaster.